Font Size: a A A

Research And Application Of Oil Field Production Index Early Warning Model Based On Artificial Neural Network

Posted on:2022-01-05Degree:MasterType:Thesis
Country:ChinaCandidate:Y Z PuFull Text:PDF
GTID:2481306323454474Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
Early warning of oilfield production indicators can help oilfield staff find abnormal production indicators in the oilfield development process in time,provide support for subsequent staff to adjust the oilfield development plan,and ensure the stable progress of oilfield development work.This paper establishes an early-warning model and system for oilfield production indicators based on artificial neural networks,and performs early-warning analysis of oilfield production indicator data of an oil production plant in Changqing Oilfield,which effectively helps oilfield managers to detect and deal with alarms in time,and reduce the loss caused by not giving early warning in time in the process of oilfield exploitation.This paper first uses the KNN algorithm to fill the oilfield production index data,uses the grey correlation analysis method and the 3? criterion to establish a suitable oilfield production early-warning indicator system,selects the GA-PSO-BP neural network to establish the oilfield production indicator early-warning model,GA-PSO-BP neural network uses GA-PSO hybrid algorithm to help BP neural network quickly find the best connection weight and threshold.GA-PSO hybrid algorithm not only maintains the rapid convergence of PSO,but also adds the global search feature of GA algorithm.By comparing the output warning experiment results of GA-PSO-BP model and PSO-BP model,GA-BP model and BP model,the GA-PSO-BP model is more accurate than other network models in forecasting changes in output fluctuations and MSE The smaller the value,the better the early warning effect,indicating the effectiveness of the GA-PSO-BP model in the early warning of oilfield production indicators.Based on the GA-PSO-BP neural network early warning model,the oilfield production indicator early warning system is designed and developed.The Java language is used for development,and the My SQL database manages the oilfield production indicator data.The oilfield production indicator early warning system realizes the management and operation of single well daily production data and monthly production data,and visually displays the output and water injection early warning results in the production indicators,helping oilfield managers to find abnormal production indicators in time for subsequent oilfield development Support for program adjustments.
Keywords/Search Tags:Oilfield early warning, Production indicators, Grey relational analysis, GA-PSO hybrid algorithm
PDF Full Text Request
Related items